Nowadays, digital systems designed to process personal data can produce harms, despite the integration of security and privacy controls. These systems are often analysed assuming they comply with requirements arising from regulations and threats deriving from external adversaries.
In real-world systems, both of these assumptions break down: implementations systematically deviate from regulatory specifications, and threats include both internal and state actors. This discrepancy gives rise to privacy harms that are not captured by current analytical tools, with impacts distributed unevenly across the affected populations.

The main objective of this PhD is to characterize and make measurable the privacy harms that emerge in privacy-preserving systems regulated under realistic threat models. This will be pursued through modeling of systems and actor ecosystems, and the construction of an operational taxonomy of harms applicable to European data spaces and digital identity systems.The final aim of the research is to develop a methodology to elicit harms and identify appropriate mitigation strategies.